AI-powered command line assistant with memory and automation capabilities for developers
AI-powered command line assistant with memory and automation capabilities for developers On Nagent, CrowTerminal is exposed as a fully-configurable developer tools integration that any agent can call — 27 actions, and API_KEY, BASIC authentication. No code is required to wire CrowTerminal into your workflow — connect it once via the External Integrations panel and reuse it across every agent you build.
Agent builders use CrowTerminal to automate the kinds of tasks developer tools teams previously handled manually. Concrete examples — each one is a single agent step in Nagent — include:
Every action and trigger is paired with a structured input/output schema (visible in the sections below), so when you wire CrowTerminal into Helix — our agentic agent builder — the editor knows exactly what each step expects and produces. Configure once, deploy anywhere across your Nagent agents.
Every operation an agent can call against CrowTerminal, with input parameters and output schema. Drop these into any step of an agent built in Helix.
CROWTERMINAL_ANALYZE_ENGAGEMENTTool to analyze engagement correlation for every field in your agent's markdown. Use when you need to understand which agent configuration fields drive engagement and get specific recommendations for improvement. Returns similarity to best/worst performing versions and field-by-field analysis.
Input parameters
The agent's current markdown/skill data to analyze. This is a dictionary containing agent configuration fields like name, description, skills, personality, etc.
The client ID to analyze engagement for
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
CROWTERMINAL_COMPARE_MDTool to compare your agent's markdown directly with all stored versions. Returns field differences showing which values differ across versions, lists missing fields not present in your current data, and provides version counts. Use when you need to understand how your current agent configuration compares to historical versions.
Input parameters
The agent's current markdown/skill data to compare against all stored versions. Contains metadata about content style, hooks, engagement, and posting schedule.
The client ID to compare markdown versions for. Used to identify which client's historical data to compare against.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
CROWTERMINAL_CREATE_WEBHOOKTool to register a new webhook for receiving real-time event notifications from CrowTerminal. Use when you need to set up asynchronous notifications for events like skill updates, data ingestion, or validation blocks.
Input parameters
The webhook URL endpoint to receive event notifications. Must be a valid HTTPS URL accessible from the internet.
Array of event types to subscribe to. Available events: 'skill.updated' (triggered when a skill is updated), 'data.ingested' (triggered when data is ingested), 'validation.blocked' (triggered when validation is blocked).
Optional secret key for webhook signature verification. If provided, this secret will be used to sign webhook payloads for security validation.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
CROWTERMINAL_DELETE_WEBHOOKTool to delete an existing webhook registration. Use when you need to remove a webhook that is no longer needed or should be replaced.
Input parameters
The unique identifier of the webhook to delete
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
CROWTERMINAL_GET_BYOK_PLATFORM_INTELTool to get algorithm insights for TikTok, Instagram, and YouTube without client-specific context. Use when you need platform intelligence data for BYOK (Bring Your Own Key) analysis workflows. This endpoint provides raw contextual algorithm data without triggering LLM inference charges.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
CROWTERMINAL_GET_CLIENT_MEMORY_CHANGELOGRetrieve human-readable change history for a client's memory. Provides a narrative view of how the client's skill data has evolved over time.
Input parameters
The unique identifier of the client whose memory changelog to retrieve.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
CROWTERMINAL_GET_CLIENT_MEMORY_PATTERNTool to track a specific field over time for trend analysis. Use when you need to understand how a particular metric evolved across versions or time periods.
Input parameters
The field name to track over time (e.g., avgEngagement). If not specified, returns patterns for all tracked fields.
The client ID to retrieve memory pattern for
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
CROWTERMINAL_GET_COMPONENTS_STATUSTool to get detailed status of each CrowTerminal service component. Returns current health status, latency, and summary statistics for all monitored components (database, cache, APIs, webhooks). Use when checking system health or diagnosing service issues.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
CROWTERMINAL_GET_DATA_TYPESTool to retrieve valid data types for ingestion across platforms. Returns available data types for TikTok, Instagram, and YouTube that can be used for data ingestion operations.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
CROWTERMINAL_GET_INCIDENTSTool to retrieve list of recent incidents from CrowTerminal with duration and affected components. Use when you need to check system status, monitor service health, or investigate recent outages or degradations.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
CROWTERMINAL_GET_PLATFORM_INTELTool to retrieve algorithm insights for TikTok, Instagram, and YouTube. Returns platform-wide intelligence about content algorithm behavior and optimization strategies. Use when you need current platform algorithm trends and recommendations.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
CROWTERMINAL_GET_SANDBOX_CLIENTTool to get mock client data for testing in the sandbox environment. Use when you need to test client-related functionality without affecting real data. No authentication required for sandbox endpoints.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
CROWTERMINAL_GET_SANDBOX_MEMORYTool to retrieve mock memory/skill data for testing purposes. Use when you need to test memory retrieval without affecting real data or requiring authentication. Part of the sandbox testing environment.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
CROWTERMINAL_GET_STATUSRetrieve CrowTerminal service status including overall health, component metrics, and uptime data. Use when you need to check the operational status of CrowTerminal services or monitor system health. No authentication required.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
CROWTERMINAL_GET_STATUS_HISTORYTool to get 7-day uptime data points ready for visualization and charting. Use when you need historical uptime metrics for monitoring dashboards or status displays.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
CROWTERMINAL_GET_UPTIMETool to retrieve historical uptime data for CrowTerminal agents. Use when you need to check system reliability, view uptime percentages for 24h/7d periods, or review recent service incidents.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
CROWTERMINAL_INGEST_BULK_DATATool to bulk ingest up to 50 analytics data points at once to CrowTerminal. Use when you need to efficiently push large amounts of platform analytics data for content creators across social media platforms. Ideal for batch uploads of retention, engagement, views, and other metrics.
Input parameters
Array of data items to ingest. Maximum 50 items per request. Each item represents analytics data for a specific creator/platform/content combination.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
CROWTERMINAL_INGEST_DATATool to ingest platform analytics data from TikTok Studio, Instagram Insights, or YouTube Analytics. Use when you need to push retention curves, demographics, traffic sources, or other engagement metrics for analysis. Supports both video-specific and channel-level data ingestion.
Input parameters
The actual analytics data payload. Structure varies by dataType. For retention: use 'curve' array with 'second' and 'retention' fields. For demographics: include age/gender/location distributions. For traffic_sources: include source names and percentages.
Optional video ID for video-specific analytics data. Required when ingesting metrics for a specific video rather than channel-level data.
The client ID for identifying the data source.
Type of analytics data being ingested. Common types include: 'retention' (audience retention curves), 'demographics' (age/gender/location data), 'traffic_sources' (referral sources), 'engagement' (likes/comments/shares), 'performance' (views/impressions).
The platform from which the analytics data originates (TikTok, Instagram, or YouTube).
Confidence level of the ingested data (0.0 to 1.0). Higher values indicate more reliable data. Useful for indicating data quality from OCR or parsing operations.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
CROWTERMINAL_LIST_WEBHOOKSTool to list all registered webhooks for the authenticated agent. Use when you need to view all webhook subscriptions and their configurations.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
CROWTERMINAL_PING_CROWTERMINALTool to check CrowTerminal service availability via a simple ping endpoint. Use when you need to verify the service is online and responding. Returns a pong confirmation with a timestamp.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
CROWTERMINAL_READ_BULK_MEMORYTool to read memory for multiple clients at once (up to 50). Use when you need to efficiently retrieve memory data for multiple creators in a single API call.
Input parameters
Array of client IDs to retrieve memory for (maximum 50 clients per request)
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
CROWTERMINAL_REGISTER_AGENTTool to self-register a new agent and obtain an API key. Use when you need to create a new agent identity in CrowTerminal. No authentication required for this endpoint. Rate limited to 5 requests per hour per IP address.
Input parameters
Name of the agent to register. Must be unique and descriptive.
Optional description of the agent's purpose or functionality.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
CROWTERMINAL_RUN_SANDBOX_ENGAGEMENT_ANALYSISTool to run mock engagement analysis in the CrowTerminal sandbox environment. Use when you need to test the engagement analysis workflow without affecting real data or when developing and validating agent configurations.
Input parameters
Agent metadata object containing configuration for mock engagement analysis.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
CROWTERMINAL_TEST_WEBHOOKTool to test a webhook URL by sending a test payload. Use when you need to verify that a webhook endpoint is properly configured and can receive requests.
Input parameters
The webhook URL to test. This URL will receive a test payload to verify it's properly configured
Optional secret for signature generation. If provided, the webhook will be signed with this secret
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
CROWTERMINAL_UPDATE_WEBHOOKTool to update an existing webhook configuration in CrowTerminal. Use when you need to modify webhook URL, change event subscriptions, or enable/disable a webhook.
Input parameters
The new webhook URL endpoint where events will be sent. Must be a valid HTTPS URL.
Array of event types to subscribe to. Common events include 'skill.updated', 'data.ingested', 'validation.blocked'.
Enable or disable the webhook. Set to true to activate, false to deactivate.
The unique identifier of the webhook to update. Must be a valid webhook ID from CrowTerminal.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
CROWTERMINAL_VALIDATE_PROPOSED_CHANGESTool to validate proposed changes against historical data before updating memory. Use when you need to check if proposed changes contradict historical patterns and receive warnings or recommendations.
Input parameters
The client ID to validate changes against
Array of proposed changes to validate against historical patterns
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
CROWTERMINAL_VALIDATE_SANDBOXTool to mock validation endpoint for testing in sandbox. Use when you need to test validation logic. Send 'tutorial' in proposedChanges to get a blocked response.
Input parameters
Array of proposed changes to validate. Send 'tutorial' as one of the values to get a blocked response for testing purposes.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
No publicly available marketplace agent is found using this tool yet. There are 47 agents privately built on Nagent that already use CrowTerminal.
Build on Nagent
Connect CrowTerminal to any Nagent agent in minutes — no API key management, no boilerplate. Just configure and deploy.
The five questions agent builders ask before adopting a new integration.
Open the External Integrations panel inside Nagent (app.nagent.ai/externalIntegration), find CrowTerminal, and click "Connect Now." You'll authenticate with API_KEY, BASIC — Nagent handles credential storage and refresh automatically. Once connected, CrowTerminal is available to any agent in your workspace.
No. Nagent provides no-code integration for every tool. Once CrowTerminal is connected, you configure its 27 actions directly in the agent builder UI — no API calls, no boilerplate, no schema management.
Helix — Nagent's agentic agent builder — lets you drop CrowTerminal steps into any workflow visually. Pick an action (e.g., one of those listed above), fill in the inputs (Helix knows the required vs. optional schema for each parameter), and connect it to upstream/downstream steps. Triggers run as the entry point of an agent, so when a CrowTerminal event fires, the agent kicks off automatically.
Every CrowTerminal action and trigger ships with a fully-typed schema — input parameters with name, type, required flag, and description, plus the output payload shape. The schemas are documented in the sections above. Helix uses these schemas to validate your configuration at build time and to type-check the data flowing between steps.
Yes. While CrowTerminal ships with 27 pre-built developer tools actions, you can layer custom logic around them inside Helix — pre/post-processing steps, conditional branches, retries, or stitching CrowTerminal together with other connected tools. For deeper customization, talk to our team about Nagent's Agentic AI Lab — forward-deployed engineers who build CrowTerminal-based workflows tailored to your business.